U.S. patent application number 11/166804 was filed with the patent office on 2006-09-14 for data relay apparatus and data relay method.
Invention is credited to Shinya Kano.
Application Number | 20060203720 11/166804 |
Document ID | / |
Family ID | 36970773 |
Filed Date | 2006-09-14 |
United States Patent
Application |
20060203720 |
Kind Code |
A1 |
Kano; Shinya |
September 14, 2006 |
Data relay apparatus and data relay method
Abstract
A failure-recovery-information storing unit stores
failure-recovery information in which a working path is associated
with a backup path. A failure-occurrence-notification receiving
unit receives a failure-occurrence notification indicating that a
failure has occurred in the working path. A backup-path searching
unit searches for a backup path corresponding to the working path
on which the failure has occurred, based on the failure-recovery
information. A path-switch processing unit carries out a
path-switch process, in such a manner that the data to be
transferred using the working path in which the failure has
occurred is transferred using the backup path.
Inventors: |
Kano; Shinya; (Kawasaki,
JP) |
Correspondence
Address: |
KATTEN MUCHIN ROSENMAN LLP
575 MADISON AVENUE
NEW YORK
NY
10022-2585
US
|
Family ID: |
36970773 |
Appl. No.: |
11/166804 |
Filed: |
June 24, 2005 |
Current U.S.
Class: |
370/228 |
Current CPC
Class: |
H04L 45/28 20130101;
H04L 45/62 20130101; H04L 41/0663 20130101; H04L 45/50 20130101;
H04L 45/22 20130101; H04L 45/00 20130101 |
Class at
Publication: |
370/228 |
International
Class: |
H04L 1/00 20060101
H04L001/00 |
Foreign Application Data
Date |
Code |
Application Number |
Mar 9, 2005 |
JP |
2005-065835 |
Claims
1. A data relay apparatus that relays data in a network in which a
failure recovery is carried out, when a failure occurs at any point
on a working path that is being used as a path through which the
data is transferred via a plurality of data relay apparatuses, by
switching the working path to a backup path for taking a detour
around a point of the failure, the data relay apparatus comprising:
a failure-recovery-information storing unit that stores
failure-recovery information in which the working path is
associated with a backup path in which the data relay apparatus
becomes a start point; a failure-occurrence-notification receiving
unit that receives a failure-occurrence notification indicating
that the failure has occurred in the working path; a backup-path
searching unit that searches for a backup path corresponding to the
working path on which the failure has occurred, based on the
failure-recovery information; and a path-switch processing unit
that carries out a path-switch process, when the backup path is
searched by the backup-path searching unit, in such a manner that
the data to be transferred using the working path in which the
failure has occurred is transferred using the backup path.
2. The data relay apparatus according to claim 1, wherein the
failure-occurrence-notification receiving unit receives the
failure-occurrence notification from a monitoring path that is
exclusively used for monitoring the failure.
3. The data relay apparatus according to claim 2, wherein a path
including a section that is shared by a plurality of working paths
is set as the monitoring path.
4. The data relay apparatus according to claim 3, wherein each of a
plurality of backup paths, for which the monitoring path is set,
has a different start point.
5. The data relay apparatus according to claim 3, wherein when each
of a plurality of backup paths corresponding to the working paths
has a different start point, a path including all of the different
start points is set as the monitoring path, and the
failure-occurrence notification is transmitted via the data relay
apparatuses on the monitoring path one by one.
6. The data relay apparatus according to claim 2, wherein a detour
tunnel is used for a plurality of backup paths having same
route.
7. The data relay apparatus according to claim 2, further
comprising a label table used for a relay of the data, wherein upon
registering information on a new working path in the label table,
the data relay apparatus registers a path that shares an existing
monitoring path, by priority.
8. The data relay apparatus according to claim 2, further
comprising a monitoring-path identifier for identifying the
monitoring path, wherein the data relay apparatus identifies the
monitoring path based on the monitoring-path identifier.
9. The data relay apparatus according to claim 1, wherein when the
failure has occurred at a point that is used by a plurality of
working paths, the failure-occurrence-notification receiving unit
receives the failure-occurrence notification from a representative
path that is used for monitoring the failure from among the working
paths.
10. The data relay apparatus according to claim 1, wherein when the
failure has occurred at a point that is used by a plurality of
working paths, the failure-occurrence-notification receiving unit
receives a plurality of failure-occurrence notifications
corresponding to the working paths, the backup-path searching unit
searches for a plurality of backup paths corresponding to the
working paths for which the failure-occurrence notifications are
received, and the path-switch processing unit excludes, when
carrying out the path-switch process, a working path for which the
path-switch process has been already carried out by any one of the
failure-occurrence notifications, from among the working paths for
which the backup paths have been searched by the backup-path
searching unit.
11. A data relay method of relaying data in a network in which a
failure recovery is carried out, when a failure occurs at any point
on a working path that is being used as a path through which the
data is transferred via a plurality of data relay apparatuses, by
switching the working path to a backup path for taking a detour
around a point of the failure, the data relay method comprising:
receiving including a start-point data-relay-apparatus receiving a
failure-occurrence notification indicating that the failure has
occurred in the working path, the start-point data relay apparatus
being a start point of a backup path corresponding to the working
path on which the failure has occurred; searching including the
start-point data relay apparatus searching for the backup path
corresponding to the working path in which the failure has
occurred, based on failure-recovery information in which the
working path is associated with the backup path; and switching
including the start-point data relay apparatus carry outing a
path-switch process, when the backup path is searched at the
searching, in such a manner that the data to be transferred using
the working path in which the failure has occurred is transferred
using the backup path.
12. The data relay method according to claim 11, wherein the
receiving includes receiving the failure-occurrence notification
from a monitoring path that is exclusively used for monitoring the
failure.
13. The data relay method according to claim 12, wherein a path
including a section that is shared by a plurality of working paths
is set as the monitoring path.
14. The data relay method according to claim 13, wherein each of a
plurality of backup paths, for which the monitoring path is set,
has a different start point.
15. The data relay method according to claim 13, wherein when each
of a plurality of backup paths corresponding to the working paths
has a different start point, a path including all of the different
start points is set as the monitoring path, and the
failure-occurrence notification is transmitted via the data relay
apparatuses on the monitoring path one by one.
16. The data relay method according to claim 12, wherein a detour
tunnel is used for a plurality of backup paths having same
route.
17. The data relay method according to claim 12, further comprising
registering information in a label table used for a relay of the
data, wherein upon registering information on a new working path in
the label table, the registering includes registering a path that
shares an existing monitoring path, by priority.
18. The data relay method according to claim 12, further comprising
identifying the monitoring path based on a monitoring-path
identifier.
19. The data relay method according to claim 11, wherein when the
failure has occurred at a point that is used by a plurality of
working paths, the receiving includes receiving the
failure-occurrence notification from a representative path that is
used for monitoring the failure from among the working paths.
20. The data relay method according to claim 11, wherein when the
failure has occurred at a point that is used by a plurality of
working paths, the receiving includes receiving a plurality of
failure-occurrence notifications corresponding to the working
paths, the searching includes searching for a plurality of backup
paths corresponding to the working paths for which the
failure-occurrence notifications are received, and the switching
includes excluding, when carrying out the path-switch process, a
working path for which the path-switch process has been already
carried out by any one of the failure-occurrence notifications,
from among the working paths for which the backup paths have been
searched at the searching.
Description
BACKGROUND OF THE INVENTION
[0001] 1. Field of the Invention
[0002] The present invention relates to a data relay apparatus and
a data relay method to relay data in a network, which carries out a
failure recovery, when a failure occurs at any point on a working
path that is being used as a path through which data is transferred
via a plurality of data relay apparatuses, by switching the path to
a backup path to make a detour to avoid the failure point, and more
particularly to a data relay apparatus and a data relay method that
can reduce the number of path used to recover the failure by
increasing the number of working paths that can be handled by a
single path used to recover the failure, such as a monitor
path.
[0003] 2. Description of the Related Art
[0004] The generalized multi-protocol label switching (GMPLS)/MPLS
is a technology to transfer data according to label information.
The label information includes a fixed-length label that is
attached to a head of a packet, a time slot in a time division
transmission, and an optical wavelength in an optical-multiplexed
transmission. In particular, a network in which data is transferred
using the fixed-length label that is attached to a head of a packet
is referred to as the MPLS. In addition, the GMPLS network uses a
single piece or a plurality of pieces of information including the
fixed-length label used in the MPLS network.
[0005] For example, in a packet transmission using the fixed-length
label, a relay node (a label switch router (LSR)) holds a label
table that represents a relation of an output label and an output
interface (IF) with respect to an input label and an input IF. At a
time of packet relay, an output IF is determined according to a
label attached to a packet received, instead of an address. Then,
the label attached to the packet is rewritten into the output label
to be relayed. By repeating this procedure, the packet is
transmitted to a destination. A relay node (start-point node) of
the MPLS network attaches the label at the beginning. The MPLS is
such kind of high-speed packet relay technology.
[0006] FIG. 34 is a schematic diagram for explaining a packet
transmission using the fixed-length label. The figure illustrates
an example of transferring a packet from LSR1 to LSR4. First of
all, the LSR1 attaches a label a to the packet to be transferred.
An LSR2 receives the packet having the label a from an IF#1,
searches for a label table, and acquires an output IF and an output
label. After that, the LSR2 rewrites the label of the packet into
an output label, and outputs the packet to an output IF. The MPLS
transfers the packet to the LSR4 of an end-point by repeating such
kind of process at each of the LSRs.
[0007] In this manner, the MPLS can enhance the speed of the packet
relay by transferring the packet according to the fixed-length
label. In addition, the MPLS can assure a bandwidth for each of the
packet flows by associating a bandwidth control in the relay node
with each of the labels.
[0008] Furthermore, in the time division transmission, each of the
nodes holds a label table that represents a relation of an output
label and an output IF with respect to an input label and an input
IF. Then, each of the nodes determines an output IF and an output
time slot according to a reception IF and a reception time slot,
and outputs data to the output time slot of the output IF. By
repeating this procedure, the data is transmitted to a
destination.
[0009] Moreover, in the optical-multiplexed transmission, each of
the nodes holds a label table that represents a relation of an
output optical wavelength and an output IF with respect to an input
optical wavelength and an input IF. Then, each of the nodes
determines an output IF and an output optical wavelength according
to a reception IF and a reception optical wavelength, converts the
reception optical wavelength into the output optical wavelength,
and outputs the output optical wavelength to the output IF. By
repeating this procedure, the data is transmitted to a
destination.
[0010] The GMPLS is a technology to carry out a data transfer in
the same mechanism by handling the fixed-length label, the time
slot, and the optical wavelength as a label.
[0011] In the GMPLS/MPLS, it is necessary to build a label table at
each of the nodes, and a path-establishing signal protocol (such as
CR-LDP/RSVP-TE) is used for building the label table. The following
is an explanation of an operation for a path establishment with the
RSVP-TE as an example.
[0012] FIG. 35 is a schematic diagram for explaining an operation
of the path-establishing signal protocol (RSVP-TE). As shown in the
figure, a start-point node that requests a path establishment
transmits a request message for the path establishment (Path
message) to an end-point node of the path by Hop-by-Hop. In the
schematic diagram shown in FIG. 35, information on a relay node to
be routed through is inserted in the Path message, for designating
a path explicitly.
[0013] The end-point node that receives the Path message returns a
path-establishment response message (Resv message) to carry out an
assignment of the label to the start-point node along the path
through which the Path message is transmitted. At this moment, a
label table for transferring data is built by registering a label
stored in the Resv message to the label table. A path ID is stored
in both of the Path message and the Resv message, and the path ID
is also registered to the label table accordingly.
[0014] In the path-establishment signal protocol, a
PathErr/ResvTear/Notify message is used as a failure message for
notifying that a failure has occurred in a path established, in
addition to the Path message indicating a request for a path
establishment and the Resv message that is a response to the Path
message.
[0015] FIG. 36 is a schematic diagram for illustrating a transfer
of a failure message using a PathErr/ResvTear message. As shown in
the figure, the PathErr/ResvTear message is transferred to the
start-point node by Hop-by-Hop along the path. On the other hand, a
Notify message is directly transferred to the start-point node that
is a destination for a message. For this reason, the Notify message
may also arrives at the start-point node along a route that is
different from the path. In the failure message, the same path ID
as the path ID stored in the Path/Resv message, so that it is
possible to carry out an association between the failure and the
path.
[0016] In the GMPLS/MPLS, when a failure occurs in a working path
for transferring data, the transfer of the data is carried out
using a backup path that makes a detour around the working path.
For example, in the MPLS, the working path is monitored using a
monitoring path, and when a failure occurs in the working path, a
failure recovery process is carried out by switching the working
path to a backup path (see, for example, Japanese Patent
Application Laid-Open Publication No. 2003-338831).
[0017] FIG. 37 is a schematic diagram for explaining a failure
recovery method using a monitoring path. As shown in the figure, in
the failure recovery using the monitoring path, a working path and
a backup path for making a traffic on the working path take a
detour at a time of a failure occurrence, using the
path-establishment signaling protocol. At this moment, the backup
path is established between the same start-point node and end-point
node as those of the working path.
[0018] Furthermore, a monitoring path for monitoring a status of
the working path is established using the path-establishment
signaling protocol. The monitoring path includes, as shown in FIG.
37, two paths in both a downstream direction and an upstream
direction along the working path. A monitoring packet is
reciprocated on the two monitoring paths periodically. This
monitoring packet is a labeled packet because it is transmitted on
the monitoring path.
[0019] When there is a failure on the working path, the start-point
node "LSR1" detects the failure by one of the following methods.
One of the methods is to detect the failure by a fact that the
monitoring packet that is supposed to be transmitted and received
periodically does not arrive within a predetermined time. Another
method is to detect the failure by a relay node on the upstream,
and to transfer a failure notification packet with a labeled packet
by the relay node.
[0020] Upon recognizing that a failure has occurred by one of the
above two methods, the start-point node transmits the traffic that
was being transmitted to the working path to the backup path to
make a detour around the failure. One monitoring path can be
established for a plurality of working paths that is established on
the same route, as shown in FIG. 37. This will decrease a total
number of the monitoring paths.
[0021] However, in the conventional technology described above,
since the backup path is limited to a plurality of paths having the
same route only, it is not possible to increase the number of the
working paths that can be handled with one monitoring path. For
this reason, the number of the monitoring paths is inevitably
increased.
[0022] Furthermore, the fact that it is necessary to establish two
monitoring paths for one sector is another factor for increasing
the number of the monitoring paths. In consequence, the number of
the monitoring paths to be managed on a network is increased,
resulting in an increase of a management load.
[0023] Moreover, since it is necessary to transmit and receive the
monitoring packet periodically, the monitoring packet consumes a
bandwidth even for a normal time. In addition, since the
notification packet for notifying the failure is inserted from the
relay node, not from an entrance node of the path, it is necessary
to install a function for inserting a packet from a relay node of
the path.
[0024] It is an object of the present invention to solve the above
problems in the conventional technology and to provide a data relay
apparatus and a data relay method that can reduce the number of
paths for a recovery of a failure by increasing the number of
working paths that can be handled by a single path for the recovery
of the failure, such as a monitoring path.
[0025] As a failure recovery method, there are several methods
disclosed so far, such as a path detour method using an error
message of the signaling protocol, a fast rerouting method explored
by IETF (Internet Draft:
draft-iert-mpls-rsvp-lsp-fastreroute-07.txt), and a path detour
method using a multiple addressing of failure information (Yasuki
fujii, Keiji Miyazaki, Kohei Iseda, "Review of preplan-type failure
recovery method", Singakugiho TM2000-60, pp. 67-72, November
2000).
[0026] However, the path detour method using the signaling protocol
has a problem that, when a failure that affects a plurality of
paths occurs, a number of message are generated, which increases a
network load. The fast rerouting method has a problem of increasing
the number of backup paths because it is necessary to establish the
same number of backup paths as the number of links through which
the working path passes. The path detour method using the multiple
addressing of the failure information has a problem that a packet
to an unrelated node is generated because a multiple-address packet
is used.
SUMMARY OF THE INVENTION
[0027] It is an object of the present invention to solve at least
the above problems in the conventional technology.
[0028] A data relay apparatus according to one aspect of the
present invention, which relays data in a network in which a
failure recovery is carried out, when a failure occurs at any point
on a working path that is being used as a path through which the
data is transferred via a plurality of data relay apparatuses, by
switching the working path to a backup path for taking a detour
around a point of the failure, includes a
failure-recovery-information storing unit that stores
failure-recovery information in which the working path is
associated with a backup path in which the data relay apparatus
becomes a start point; a failure-occurrence-notification receiving
unit that receives a failure-occurrence notification indicating
that the failure has occurred in the working path; a backup-path
searching unit that searches for a backup path corresponding to the
working path on which the failure has occurred, based on the
failure-recovery information; and a path-switch processing unit
that carries out a path-switch process, when the backup path is
searched by the backup-path searching unit, in such a manner that
the data to be transferred using the working path in which the
failure has occurred is transferred using the backup path.
[0029] A data relay method according to another aspect of the
present invention, which is for relaying data in a network in which
a failure recovery is carried out, when a failure occurs at any
point on a working path that is being used as a path through which
the data is transferred via a plurality of data relay apparatuses,
by switching the working path to a backup path for taking a detour
around a point of the failure, includes receiving including a
start-point data-relay-apparatus receiving a failure-occurrence
notification indicating that the failure has occurred in the
working path, the start-point data relay apparatus being a start
point of a backup path corresponding to the working path on which
the failure has occurred; searching including the start-point data
relay apparatus searching for the backup path corresponding to the
working path in which the failure has occurred, based on
failure-recovery information in which the working path is
associated with the backup path; and switching including the
start-point data relay apparatus carry outing a path-switch
process, when the backup path is searched at the searching, in such
a manner that the data to be transferred using the working path in
which the failure has occurred is transferred using the
backup-path.
[0030] The above and other objects, features, advantages and
technical and industrial significance of this invention will be
better understood by reading the following detailed description of
presently preferred embodiments of the invention, when considered
in connection with the accompanying drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
[0031] FIG. 1 is a schematic diagram of a data relay apparatus
according to an embodiment of the present invention;
[0032] FIG. 2 is an example of a relay-label table;
[0033] FIG. 3 is an example of a failure-recovery table;
[0034] FIG. 4 is a schematic diagram for illustrating an
establishment of a working path;
[0035] FIG. 5 is a schematic diagram for illustrating an
establishment of a backup path;
[0036] FIG. 6 is a schematic diagram for illustrating an
establishment of a monitoring path;
[0037] FIG. 7 is a first schematic diagram for illustrating a
correspondence between a working path, a backup path, and a
monitoring path;
[0038] FIG. 8 is a first schematic diagram for illustrating a
transmission of an error message at a time of a failure
occurrence;
[0039] FIG. 9 is a first schematic diagram for illustrating a path
switch for taking a detour around a failure;
[0040] FIG. 10 is a first schematic diagram for illustrating an
establishment of a working path, a backup path, and a monitoring
path;
[0041] FIG. 11 is a second schematic diagram for illustrating a
correspondence between a working path, a backup path, and a
monitoring path;
[0042] FIG. 12 is a second schematic diagram for illustrating a
transmission of an error message at a time of a failure
occurrence;
[0043] FIG. 13 is a second schematic diagram for illustrating a
path switch for taking a detour around a failure;
[0044] FIG. 14 is a schematic diagram for illustrating an
establishment of a working path, a Bypass tunnel, and a monitoring
path;
[0045] FIG. 15 is a schematic diagram for illustrating a
correspondence between a working path, a Bypass tunnel, and a
monitoring path;
[0046] FIG. 16 is a third schematic diagram for illustrating a
transmission of an error message at a time of a failure
occurrence;
[0047] FIG. 17 is a third schematic diagram for illustrating a path
switch for taking a detour around a failure;
[0048] FIG. 18 is a schematic diagram for illustrating a monitoring
path corresponding to a plurality of backup paths in different
detour sections (part 1);
[0049] FIG. 19 is a schematic diagram for illustrating a monitoring
path corresponding to a plurality of backup paths in different
detour sections (part 2);
[0050] FIG. 20 is a schematic diagram for illustrating established
working path, backup path, and monitoring path;
[0051] FIG. 21 is a schematic diagram for illustrating an addition
of a working path;
[0052] FIG. 22 is a third schematic diagram for illustrating a
correspondence between a working path, a backup path, and a
monitoring path;
[0053] FIG. 23 is a first schematic diagram for illustrating an
establishment of a working path and a backup path;
[0054] FIG. 24 is a schematic diagram for illustrating an
establishment of a working path, a backup path, and a
representative path;
[0055] FIG. 25 is an example of a failure-recovery table when a
representative path is used;
[0056] FIG. 26 is a fourth schematic diagram for illustrating a
transmission of an error message at a time of a failure
occurrence;
[0057] FIG. 27 is a fourth schematic diagram for illustrating a
path switch for taking a detour around a failure;
[0058] FIG. 28 is a second schematic diagram for illustrating an
establishment of a working path and a backup path;
[0059] FIG. 29 is an example of a failure-recovery table when
switching a plurality of paths with a single error message;
[0060] FIG. 30 is a schematic diagram for illustrating a
correspondence between a working path and a backup path;
[0061] FIG. 31 is a fifth schematic diagram for illustrating a
transmission of an error message at a time of a failure
occurrence;
[0062] FIG. 32 is a fifth schematic diagram for illustrating a path
switch for taking a detour around a failure;
[0063] FIG. 33 is another example of a failure-recovery table when
switching a plurality of paths with a single error message;
[0064] FIG. 34 is a schematic diagram for illustrating a
transmission of a packet by using a fixed-length label;
[0065] FIG. 35 is a schematic diagram for illustrating an operation
of a path-establishment signaling protocol (RSVP-TE);
[0066] FIG. 36 is a schematic diagram for illustrating a
transmission of an error message by using a PathErr/ResvTear
message; and
[0067] FIG. 37 is a schematic diagram for illustrating a failure
recovery method using a monitoring path.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0068] Exemplary embodiments of a data relay apparatus and a data
relay method according to the present invention will be explained
in detail below with reference to the accompanying drawings.
[0069] FIG. 1 is a schematic diagram of a data relay apparatus
according to an embodiment of the present invention. As shown in
the figure, the data relay apparatus 100 includes a data-receiving
unit 110, a data transmitting unit 120, a data relay unit 130, a
relay-label table 140a, a backup-path-label table 140b, a
monitor-path label table 140c, a control-packet receiving unit 150,
a control-packet transmitting unit 160, a failure-recovery table
170, a working path control unit 180a, a backup-path control unit
180b, a monitor-path control unit 180c, and a path-switch
processing unit 190.
[0070] The data-receiving unit 110 receives data from a network,
and delivers the data received to the data relay unit 130. The data
transmitting unit 120 receives the data from the data relay unit
130, and transmits the data received to the network.
[0071] The data relay unit 130 relays the data based on the
relay-label table 140a. In other words, the data relay unit 130
receives the data and an input interface (IF) from the
data-receiving unit 110, and searches for an output label and an
output IF from an input label and the input IF based on the
relay-label table 140a. Then, the data relay unit 130 replaces the
input label with the output label, and delivers the data to the
data transmitting unit 120 by designating the output IF.
[0072] The relay-label table 140a stores label information used for
relaying the data. FIG. 2 is an example of the relay-label table
140a. As shown in the figure, the relay-label table 140a stores a
path ID for identifying a path, the input IF, the input label, the
output IF, and the output label, for each of the working paths.
[0073] The backup-path-label table 140b stores label information
for each of the backup paths, having the same data structure as
that of the relay-label table 140a. The monitor-path label table
140c stores label information for each of the monitor paths, having
the same data structure as that of the relay-label table 140a.
[0074] The control-packet receiving unit 150 receives a control
packet from the network, and delivers the control packet received
to any one of the working path control unit 180a, the backup-path
control unit 180b, and the monitor-path control unit 180c, based on
information on a path of the control packet received.
[0075] The control packet includes a path-establishment request
message (PATH message), a path-establishment response message (RESV
message), and an error message. The path-establishment request
message (PATH message) is a message for requesting a path
establishment for the working path, the backup path, and the
monitoring path, including the path ID.
[0076] The path-establishment response message (RESV message) is a
response message to the path-establishment request message for the
working path, the backup path, and the monitoring path, including
the path ID and a label for assignment.
[0077] The error message is a message for notifying an occurrence
of an error on a path to a start-point node, including information
on a point at which the error has occurred and the path ID. The
error message is delivered to the monitor-path control unit
180c.
[0078] The control-packet transmitting unit 160 receives the
control packet from the working path control unit 180a, the
backup-path control unit 180b, and the monitor-path control unit
180c, and transmits the control packet received to the network.
[0079] The failure-recovery table 170 is used by the path-switch
processing unit 190, when a failure occurs on the working path, to
switch the working path to the backup path. FIG. 3 is an example of
a failure-recovery table 170. As shown in the figure, the
failure-recovery table 170 stores a working path ID for identifying
the working path, a backup-path ID for identifying the backup path,
and a monitoring-path ID for identifying the monitoring path,
corresponding to the backup path in which the data relay apparatus
itself is a start point.
[0080] The failure-recovery table 170 can make a start point of the
backup path and a start point of the monitoring path different as
long as the start point of the backup path is included in the
monitoring path, by storing the working path and the monitoring
path in association with the backup path in which the data relay
apparatus itself is the start point.
[0081] The working path control unit 180a processes a control
packet regarding to the working path. In other words, the working
path control unit 180a transfers a PATH message following
passing-node information included in the PATH message requesting a
path establishment of the working path, and stores a transmission
node of the PATH message.
[0082] When an RESV message is received, label information included
in the RESV message received is assigned as the output label with a
reception interface as the output IF; label information included in
the PATH message is assigned as the input label with an interface
through which the PATH message is received as the input IF.
[0083] Subsequently, the input label is replaced with a label of
the RESV message received, and transmitted to the transmission node
of the PATH message. The interface and the label information
assigned are registered in the relay-label table 140a, and a path
ID is registered in the working path ID of the failure-recovery
table 170.
[0084] The backup-path control unit 180b carries out the same
process, as the working path control unit 180a carries out for the
control packet regarding to the working path, to the control packet
regarding to the backup path, and registers the interface and the
label information assigned in the backup-path-label table 140b
instead of the relay-label table 140a.
[0085] In addition, when the data relay apparatus itself is a start
point of the backup path, the backup-path control unit 180b
registers a backup-path ID corresponding to the working path ID
registered in the failure-recovery table 170, and when the data
relay apparatus itself is not the start point of the backup path,
deletes the working path ID registered in the failure-recovery
table 170 corresponding to the backup-path ID.
[0086] The monitor-path control unit 180c carries out the same
process, as the working path control unit 180a carries out for the
control packet regarding to the working path, to the control packet
regarding to the monitoring path, and registers the interface and
the label information assigned in the monitoring-path-label table
140c instead of the relay-label table 140a.
[0087] In addition, when the data relay apparatus itself is a start
point of the backup path, the monitoring-path control unit 180c
registers a monitoring-path ID corresponding to the working path ID
and the backup-path ID registered in the failure-recovery table
170. Furthermore, upon receiving an error message, the
monitoring-path control unit 180c delivers the error message to the
path-switch processing unit 190 notifying that an error has
occurred on a path.
[0088] The path-switch processing unit 190, upon receiving a
notification from the monitor-path control unit 180c of a reception
of the error message, carries out a path switch from the working
path to the backup path. In other words, the path-switch processing
unit 190 determines whether the monitoring-path ID stored included
in the error message is registered by referring to the
failure-recovery table 170. When the monitoring-path ID is
registered, the path-switch processing unit 190 specifies the
working path and the backup path corresponding to the
monitoring-path ID, because the data relay apparatus itself is the
start point of the backup path. Then, the path-switch processing
unit 190 extracts information on the backup path from the
backup-path-label table 140b, and registers the information
extracted in the relay-label table 140a. With this mechanism, the
traffic on the working path is transferred to the backup path.
[0089] In this manner, the path-switch processing unit 190
determines whether the monitoring-path ID stored included in the
error message is registered in the failure-recovery table 170,
specifies the working path and the backup path corresponding to the
monitoring-path ID when the monitoring-path ID is registered, and
carries out a path switch so that the traffic on the working path
is transferred to the backup path, making it possible to recover a
failure of the working path.
[0090] Following explanations are for several cases of operations
of the failure recovery by the data relay apparatus 100 according
to the present embodiment. As the first case a failure-recovery
operation by the data relay apparatus 100 according to the present
embodiment is explained with reference to FIGS. 4 to 9 for a
network having seven nodes. The node means the data relay
apparatus. Hereinafter, the relay-label table 140a, the
backup-path-label table 140b, and the monitoring-path-label table
140c are collectively referred to as a label table.
[0091] First, a working path is established with a route of "node
1".fwdarw."node 2".fwdarw."node 3".fwdarw."node 4".fwdarw."node 5",
by using the path-establishment signaling protocol (RSVP-TE) (FIG.
4). This causes the same table as a label table belonging to the
"node 2", as shown in FIG. 4, to be built on a whole route of the
working path.
[0092] To recover from a failure occurred at "node 2"-"node
3"-"node 4", a backup path of a route "node 2".fwdarw."node
6".fwdarw."node 7".fwdarw."node 4" with the "node 2" as a start
point and the "node 4" as an end point is established. This backup
path is also established by using the path-establishment signaling
protocol (RSVP-TE), like the working path (FIG. 5). This causes the
same table as a label table belonging to the "node 2" and "node 6",
as shown in FIG. 5, to be built on a whole route of the backup
path.
[0093] The "node 2" and "node 4" are registered in combination with
a label and interface information of a corresponding working path,
based on association information with the working path included in
the PATH message of the backup path. As the association information
with the working path, a method of determining a path ID according
to an existing association rule, a method of storing the working
path information in the PATH message, and the like can be used. In
the label table of the "node 2" shown in FIG. 5 includes the input
label and the input interface information of the working path
1.
[0094] In addition, the "node 2" that becomes a path-switch node at
the time of a failure holds the failure-recovery table 170 that is
an association table of the working path and the backup path. In
this example, the working path ID is "#1", and the backup-path ID
is "#10".
[0095] In order to monitor a failure occurs between the "node
2"-"node 3"-"node 4" through which the working path passes, a
monitoring path is established on the same route (FIG. 6). To
establish the monitoring path, the path-establishment signaling
protocol (RSVP-TE) is used.
[0096] Furthermore, to indicate that the path established is the
monitoring path, when the path-establishment signaling protocol is
carried out, information, such as a flag or an object indicating
that the path is the monitoring path, is implanted, or a path ID
indicating that the path is the monitoring path is assigned.
[0097] This causes the same table as the label table held by the
"node 2" shown in FIG. 6 to be built on the whole route of the
monitoring path. In the label table of the "node 2" shown in FIG.
6, the input label and the interface information are not registered
because the "node 2" is a start-point node of the monitoring
path.
[0098] The "node 2" that becomes a path-switch node at the time of
a failure holds the failure-recovery table 170 that is an
association table of the working path and the monitoring path (FIG.
7). In this example, the monitoring-path ID is "#100".
[0099] As explained above, in the path-establishment signaling
protocol, there is an error message that notifies that an error has
occurred on the path established, in addition to the Path message
indicating a path establishment request and the Resv message that
is a response to the request. A PathErr/ResvTear/Notify message is
corresponding to the error message. This is used according to a
type of the error, however, with any kind of error message, it is
possible to realize the following operation. In the error message,
a path ID that is stored in the Path/Resv message at the time of
path establishment is stored.
[0100] When a failure occurs between the "node 2"-"node 3"-"node
4", a relay node at an upstream detects the failure. In this
example, there is a failure between the "node 3"-"node 4". Then,
the "node 3" that detected the failure transmits an error message
with respect to the monitoring path to the "node 2", to notify that
the failure has occurred to the path-switch node "node 2" (FIG. 8).
In the error message with respect to the monitoring path, a
monitoring-path ID is included.
[0101] The "node 2" that received the error message searches the
failure-recovery table 170, based on the path ID stored in the
error message, and identifies the working path and the backup path.
Then, the failure is recovered by transmitting the traffic of the
working path searched to a corresponding backup path (FIG. 9).
[0102] In the first case, when a failure occurs between the "node
3"-"node 4", the "node 3" detects the failure, and transmits an
error message with respect to a monitoring path to the "node 2"
without transmitting the error message with respect to the working
path. In this manner, by suppressing a function of transmitting the
error message defined by the path-establishment signaling protocol,
it is possible to reduce an unnecessary message.
[0103] Furthermore, in the first case, although a great effect is
not obtained because one monitoring path is associated with one
working path, if one monitoring path is associated with a plurality
of working paths, as will be explained in a second case, it is
possible to greatly reduce the number of error messages created at
a time of a failure.
[0104] In addition, in the first case, when a failure occurs
between the "node 3"-"node 4", the "node 3" detects the failure,
and transmits an error message with respect to a monitoring path to
the "node 2". However, a transmission of the error message with
respect to the working path is defined in the path-establishment
signaling protocol. Therefore, it is necessary to add a function of
suppressing the transmission of the error message with respect to
the working path.
[0105] On the other hand, it is possible to realize the above
operation without adding the function of suppressing the
transmission of the error message with respect to the working path,
by transmitting the error message with respect to the monitoring
path prior to transmitting the error message with respect to the
working path. In other words, when two packets are transmitted
substantially at the same time, there is a possibility that the
packet arrived later is discarded when being relayed due to a load
and a process capability of a relay node. Therefore, by
transmitting the error packet for the monitoring path in advance,
it is possible to avoid discarding of the error packet for the
monitoring path, so that the error packet for the monitoring path
arrives at a path-switch node.
[0106] In the first case, the working path is established first,
and then the backup path is established followed by the monitoring
path. However, the order of establishment of the backup path and
the monitoring path can be reversed, so that the monitoring path is
established before the backup path. In the first case in which the
backup path is established before the monitoring path, if the
failure occurs before the monitoring path is established, there is
a possibility that the traffic cannot be detoured from the working
path to the backup path. However, if the monitoring path is
established before the backup path, it is possible to detour the
traffic from the working path to the backup path no matter when the
failure occurs after the backup path is established.
[0107] Moreover, in the first case, the backup path is established
only at a part of the working path, and the traffic is detoured
from the working path to the backup path when the failure occurs at
the part of the working path. However, it is not necessary to limit
the establishment of the backup path only at the part of the
working path. Instead, it is possible to establish the backup path
at other part of the working path. In general, a plurality of
backup paths is established to avoid a failure even when the
failure occurs at any point on the working path.
[0108] When the error message is delivered to the start-point node
of the monitoring path, if the PathErr/ResvTear message is used as
the error message, the error message is transferred on the
monitoring path by Hop-by-Hop. On the other hand, if the Notify
message is used as the error message, the error message is directly
transferred to the start-point node that is a destination for the
packet. For this reason, the error message may be transferred
through different path from the monitoring path.
[0109] In addition, the label table and the failure-recovery table
are built in each of the apparatuses manually, without using the
path-establishment signaling protocol to establish the working
path, the backup path, and the monitoring path.
[0110] As the second case, a failure-recovery operation by the data
relay apparatus 100 according to the present embodiment for a
network configured with nine nodes will be explained with reference
to FIGS. 10 to 13. First, a "working path 1" having a route of
"node 1".fwdarw."node 2".fwdarw."node 3".fwdarw."node
4".fwdarw."node 5" and a "working path 2" having a route of "node
8".fwdarw."node 2".fwdarw."node 3".fwdarw."node 4".fwdarw."node 9"
are established using the path-establishment signaling protocol
(RSVP-TE).
[0111] To recover from a failure occurred at "node 2"-"node
3"-"node 4", a "backup path 1" and a "backup path 2" having a route
of "node 2".fwdarw."node 6".fwdarw."node 7".fwdarw."node 4" with
the "node 2" as a start point and the "node 4" as an end point are
established. The "backup path 1" is used as a detour for the
"working path 1", and the "backup path 2" is used as a detour for
the "working path 2". The "backup path 1" and the "backup path 2"
are established using the path-establishment signaling protocol
(RSVP-TE), in the same manner as the working path.
[0112] The "node 2" that becomes a path-switch node at the time of
a failure holds the failure-recovery table 170. In this example,
the path IDs of the "working path 1" and the "working path 2" are
"#1" and "#2", respectively, and the path IDs of the "backup path
1" and the "backup path 2" are "#10" and "#20", respectively.
[0113] The "working path 1" and the "working path 2" pass the same
section ("node 2"-"node 3"-"node 4"). In order to monitor a failure
occurs in this section, a monitoring path is established on the
same route. To establish the monitoring path, the
path-establishment signaling protocol (RSVP-TE) is used.
[0114] The "working path 1", the "working path 2", the "backup path
1", "backup path 2", and the monitoring path are shown in FIG. 10.
An example of the label table built with the path-establishment
signaling protocol is also shown in FIG. 10.
[0115] The "node 2" that becomes a path-switch node at the time of
a failure holds the failure-recovery table 170. In this example,
when the path ID of the monitoring path is "#100", the
failure-recovery table 170 becomes the one shown in FIG. 11.
[0116] When a failure occurs between the "node 2"-"node 3"- "node
4", a data relay apparatus on an upstream side detects the failure.
In this example, the failure is occurs between the "node 3"-"node
4". The "node 3" that detected the failure transmits an error
message with respect to a monitoring path to the "node 2" to notify
the path-switch "node 2" that the failure has occurred. A path ID
of the monitoring-path is stored in the error message with respect
to the monitoring path (FIG. 12).
[0117] The "node 2" receives the error message, searches the
failure-recovery table 170 based on the path ID stored n the error
message, and identifies the working path and the backup path. In
the second case, two sets of the working path and the backup path
are obtained. Then, the "node 2" recovers from the failure by
transferring the traffic on the working path obtained to the
corresponding backup path. As a result, it is possible to recover
two working paths affected from the failure with a reception of the
error message with respect to one monitoring path (FIG. 13).
[0118] In this manner, by using the data relay apparatus 100
according to the present embodiment, it is possible to set one
monitoring path for a plurality of working paths, and to reduce the
number of monitoring paths. Furthermore, it is possible to reduce
the error message created at the time of a failure.
[0119] Subsequently, as the third case, a failure-recovery
operation by the data relay apparatus 100 according to the present
embodiment for a network configured with nine nodes will be
explained with reference to FIGS. 14 to 17. First, a "working path
1" having a route of "node 1".fwdarw."node 2".fwdarw."node
3".fwdarw."node 4".fwdarw."node 5" and a "working path 2" having a
route of "node 8".fwdarw."node 2".fwdarw."node 3".fwdarw."node
4".fwdarw."node 9" are established using the path-establishment
signaling protocol (RSVP-TE).
[0120] To recover from a failure occurred at "node 2"-"node
3"-"node 4", a Bypass tunnel having a route of "node
2".fwdarw."node 6".fwdarw."node 7".fwdarw."node 4" with the "node
2" as a start point and the "node 4" as an end point is
established. The Bypass tunnel is established using the
path-establishment signaling protocol (RSVP-TE), in the same manner
as the working path.
[0121] At the time of a failure, the route of the working path is
replaced with the Bypass tunnel. In addition, the traffic
transferred on the Bypass tunnel includes a Bypass-tunnel label
attached in front of the working path label. The "node 2" attaches
the label used by the "node 4" before the failure to the traffic as
a current label, so that the "node 4" can transfer the traffic
transferred through the Bypass tunnel to the working path.
[0122] In addition, the "node 2" that becomes a path-switch node at
the time of a failure holds the failure-recovery table 170 that is
a table for associating the working path and the backup path. In
this example, the path IDs of the "working path 1" and the "working
path 2" are "#1" and "#2", respectively, and the path ID of the
Bypass tunnel is "#1000".
[0123] The "working path 1" and the "working path 2" pass the same
section ("node 2"-"node 3"-"node 4"). In order to monitor a failure
occurs in this section, a monitoring path is established on the
same route. To establish the monitoring path, the
path-establishment signaling protocol (RSVP-TE) is used.
[0124] Furthermore, to indicate that the path established is the
monitoring path, when the path-establishment signaling protocol is
carried out, information, such as a flag or an object indicating
that the path is the monitoring path, is implanted, or a path ID
indicating that the path is the monitoring path is assigned.
[0125] The "working path 1", the "working path 2", the Bypass
tunnel, and the monitoring path established are shown in FIG. 14.
In addition, an example of the label table built with the
path-establishment signaling protocol is also shown in FIG. 14.
[0126] The "node 2" that becomes a path-switch node at the time of
a failure holds the failure-recovery table 170 that is a table for
associating the Bypass tunnel and the monitoring path (FIG. 15). In
this example, the path ID of the monitoring path is "#100".
[0127] When a failure occurs between the "node 2"-"node 3"- "node
4", a relay node at an upstream detects the failure. In this
example, there is a failure between the "node 3"- "node 4". Then,
the "node 3" that detected the failure transmits an error message
with respect to the monitoring path to the "node 2", to notify that
the failure has occurred to the path-switch node "node 2" (FIG.
16). In the error message with respect to the monitoring path, a
monitoring-path ID is included.
[0128] The "node 2" that received the error message searches the
failure-recovery table 170, based on the path ID stored in the
error message, and identifies the working path and the Bypass
tunnel. In the third case, two sets of the working path and the
Bypass tunnel are obtained. Then, the working path obtained is
replaced with the Bypass tunnel.
[0129] An output label used at the time of a failure is attached to
the traffic. The traffic is further attached with a Bypass-tunnel
label, and transferred to the Bypass tunnel. With this mechanism,
it is possible to recover from the failure (FIG. 17). As a result,
it is possible to recover two working paths affected from the
failure with a reception of the error message with respect to one
monitoring path.
[0130] In this manner, by using the data relay apparatus 100
according to the present embodiment, it is possible to set one
monitoring path for a plurality of working paths, and to reduce the
number of monitoring paths. Furthermore, it is possible to reduce
the error message created at the time of a failure.
[0131] In the explanation so far, a case in which the start point
and the end point of the backup path are the same as those of the
monitoring path has been explained. However, it is also possible to
set the start point and the end point of the backup path and the
start point and the end point of the monitoring path differently.
As the fourth case, the case in which the end point of the backup
path is different from the end point of the monitoring path is
explained with reference to FIG. 18.
[0132] First, a "working path 1" having a route of "node
1".fwdarw."node 2".fwdarw."node 3".fwdarw."node 4".fwdarw."node
13".fwdarw."node 14" and a "working path 2" having a route of "node
12".fwdarw."node 2".fwdarw."node 3".fwdarw."node 4".fwdarw."node
5".fwdarw."node 6" are established (FIG. 18).
[0133] The "working path 1" and the "working path 2" pass the same
section "node 2"-"node 3"-"node 4". A backup path to take a detour
around a failure occurred in this section is established. "The
backup path 1" corresponding to the "working path 1" is established
with a route "node 2".fwdarw."node 7".fwdarw."node 8".fwdarw."node
13", and the "backup path 2" corresponding to the "working path 2"
is established with a route "node 2".fwdarw."node 9".fwdarw."node
10".fwdarw."node 11".fwdarw."node 5".
[0134] Furthermore, in the fourth case, because it is not necessary
to make the end point of the backup path and the end point of the
monitoring path the same, a monitoring path is established for
monitoring a failure occurs at the section "node 2"-"node 3"-"node
4" through which both the "working path 1" and the "working path 2"
pass.
[0135] The "node 2" registers the monitoring path, the "backup path
1", and the "backup path 2" in the failure-recovery table 170, in
association with each other. In this example, the path IDs of the
"working path 1" and the "working path 2" are "#1" and "#2",
respectively, and the path IDs of the "backup path 1" and the
"backup path 2" are "#10" and "#20", respectively. The path ID of
the monitoring path is "#100".
[0136] When a failure occurs between the "node 2"-"node 3"-"node
4", a data relay apparatus on an upstream side detects the failure.
In this example, the failure is occurs between the "node 3"-"node
4". The "node 3" that detected the failure transmits an error
message with respect to a monitoring path to the "node 2" to notify
the path-switch "node 2" that the failure has occurred. A path ID
of the monitoring-path is stored in the error message with respect
to the monitoring path.
[0137] The "node 2" receives the error message, searches the
failure-recovery table 170 based on the path ID stored in the error
message, and identifies the working path and the backup path. In
the fourth case, two sets of the working path and the backup path
are obtained. Then, the "node 2" recovers from the failure by
transferring the traffic on the working path obtained to the
corresponding backup path. As a result, it is possible to recover
two working paths affected from the failure with a reception of the
error message with respect to one monitoring path.
[0138] In this manner, by using the data relay apparatus 100
according to the present embodiment, it is possible to manage a
plurality of backup paths to take a detour around different
sections with one monitoring path, and to reduce the number of
monitoring paths.
[0139] In the fourth case, the backup path is established with
respect to a part of sections of the working path, and the traffic
is detoured only when a failure occurs at the part of sections.
However, it is not limited to establish the backup path only for
the part of sections, but the backup path can be established for
other sections according to a necessity. In other words, in the
above explanation, a part of the network is picked up to make the
explanation simple.
[0140] Although the detour path that can be set regarding to the
"working path 1" is "node 2"-"node 3"-"node 4"- "node 13" in the
fourth case, another detour path can be established for other
sections in an actual network. For example, when there is a link
between "node 14"-"node 5", a "backup path 1'" can be established
to take a detour around "node 4"-"node 13"-"node 14".
[0141] In this kind of case, it is possible to establish a
monitoring path between the "node 4"-"node 13"-"node 14", and apply
a similar failure-recovery mechanism. In this manner, by making it
possible to establish a detour path with respect to all links of
the working path, it is possible to protect the whole working path
from the failure.
[0142] In the explanation so far, a method of establishing
monitoring path at a common section of a plurality of working paths
and associating the monitoring path with the backup path has been
explained. However, it is also possible to establish a monitoring
path at other sections than the common section of the working
paths. As the fifth case, a method of reducing the number of the
monitoring paths by associating the monitoring path established at
a section bigger than the common section in a direction of the
upstream with the backup path is explained with reference to FIG.
19.
[0143] First, a "working path 1" having a route of "node
12".fwdarw."node 3".fwdarw."node 4".fwdarw."node 5".fwdarw."node
13" and a "working path 2" having a route of "node 1".fwdarw."node
2".fwdarw."node 3".fwdarw."node 4".fwdarw."node 5".fwdarw."node 6"
are established. It is assumed that a "backup path 1" and a "backup
path 2" corresponding to the "working path 1" and the "working path
2", respectively, are established.
[0144] At this moment, a monitoring path including both sections
around which the "backup path 1" and the "backup path 2" take a
detour is established at "node 2".fwdarw."node 3".fwdarw."node
4".fwdarw."node 5". The "node 2".fwdarw."node 3".fwdarw."node
4".fwdarw."node 5" where the monitoring path is established is a
bigger section than the "working path 1" and the "working path 2"
in a direction of the upstream side. For this reason, a start-point
node of the monitoring path and start-point nodes of all of the
backup paths are not the same.
[0145] In this example, the path IDs of the "working path 1" and
the "working path 2" are "#1" and "#2", respectively, and the path
IDs of the "backup path 1" and the "backup path 2" are "#10" and
"#20", respectively. The path ID of the monitoring path is "#100".
The "node 2" that is a start-point node of the "backup path 2"
associates the "backup path 2" with the monitoring path, and
registers a result of the association in the failure-recovery table
170. Furthermore, the "node 3" that is a start-point node of the
"backup path 1" associates the "backup path 1" with the monitoring
path, and registers a result of the association in the
failure-recovery table 170.
[0146] When a failure occurs between the "node 2"-"node 3"- "node
4"-"node 5", a data relay apparatus on an upstream side detects the
failure. In this example, the failure is occurs between the "node
4"-"node 5". The "node 4" that detected the failure transmits an
error message with respect to a monitoring path to the "node 2" to
notify the path-switch "node 2" that the failure has occurred. In
this example, a PathErr/Resv err message is used for the error
message.
[0147] In this case, because the error message is transferred on
the monitoring path by Hop-by-Hop, all the nodes on the monitoring
path on a side of the upstream than the transmission node of the
error message (in this case, the "node 3" that is the relay node
and the "node 2" that is the start-point node) become to be able to
handle the error message.
[0148] The "node 2" and the "node 3" that relayed the error message
searches the failure-recovery table 170 based on the path ID stored
in the error message, and identifies the working path and the
backup path. The "working path 2" and the "backup path 2" are
obtained from the "node 2", and the "working path 1" and the
"backup path 1" are obtained from the "node 3". Then, a recovery
from the failure is performed by transferring the traffic on the
working path obtained to the corresponding backup path.
[0149] In this manner, by using the data relay apparatus 100
according to the present embodiment, it is possible to detour the
traffic of a plurality of backup paths having different sections
with an error message with respect to one monitoring path, and to
reduce the number of monitoring paths.
[0150] As the sixth case, a case in which a path from the "node 1"
to the "node 5" is newly added when a set of the "working path 2"
and the "backup path 2", and a monitoring path are already
established, as shown in FIG. 20, will be explained. When the "node
1" calculates a route from the "node 1" to the "node 5", it
calculates a path that reduces the number of monitoring paths in
the whole network, by sharing the monitoring path already
established.
[0151] In the example shown in FIG. 20, although the shortest route
is "node 1".fwdarw."node 6".fwdarw."node 7".fwdarw."node 5", the
"node 1" selects "node 1".fwdarw."node 2".fwdarw."node
3".fwdarw."node 4".fwdarw."node 5" (FIG. 21). Then, the "node 1"
establishes a "working path 1" having a route of "node
1".fwdarw."node 2".fwdarw."node 3".fwdarw."node 4".fwdarw."node 5",
using the path-establishment signaling protocol (RSVP-TE).
[0152] At this moment, upon recognizing that the "working path 1"
passes the same route as the monitoring path already established,
the "node 2" establishes a "backup path 1" from the "node 2" that
is a start-point node of the monitoring path to the "node 4" that
is and an end-point node of the monitoring path at a route of "node
2".fwdarw."node 6".fwdarw."node 7".fwdarw."node 4". The "backup
path 1" corresponding to the "working path 1", and the monitoring
path are registered in the failure-recovery table 170.
[0153] In this example, the path IDs of the "working path 1" and
the "working path 2" are "#1" and "#2", respectively, and the path
IDs of the "backup path 1" and the "backup path 2" are "#10" and
"#20", respectively. If the path ID of the monitoring path is
"#100", the failure-recovery table 170 becomes the one shown in
FIG. 22.
[0154] In this manner, when selecting a route, it is possible to
reduce the number of monitoring paths in the whole network, by
selecting a route that shares the same monitoring path. In
addition, because the association of the working path, the backup
path, and the monitoring path is carried out automatically at the
time of establishing a path, it is possible to complete a setting
of the path in a short time.
[0155] So far, a case in which the error message is transferred
using the monitoring path, however, it is also possible to recover
a failure of a plurality of paths with a single error message. As
the seventh case, an operation of a failure recovery without using
a monitoring path is explained for a network having nine nodes,
with reference to FIGS. 23 to 27.
[0156] First, two working paths are established by using the
path-establishment signaling protocol (RSVP-TE). One of the two
paths is a "working path 1" with a route of "node 1".fwdarw."node
2".fwdarw."node 3".fwdarw."node 4".fwdarw."node 5"; and the other
is a "working path 2" with a route of "node 8".fwdarw."node
2".fwdarw."node 3".fwdarw."node 4".fwdarw."node 9".
[0157] To recover from a failure occurred at "node 2"-"node
3"-"node 4", a "backup path 1" and a "backup path 2" with a route
"node 2".fwdarw."node 6".fwdarw."node 7".fwdarw."node 4" with the
"node 2" as a start point and the "node 4" as an end point are
established. The "backup path 1" is used as a detour for the
"working path 1", and the "backup path 2" is used as a detour for
the "working path 2". The "backup path 1" and the "backup path 2"
are also established by using the path-establishment signaling
protocol (RSVP-TE), like the working path.
[0158] The "working path 1", the "working path 2", the "backup path
1", "backup path 2" established are shown in FIG. 23. An example of
the label table built with the path-establishment signaling
protocol is also shown in FIG. 23.
[0159] The "working path 1" and the "working path 2" pass the same
section ("node 2"-"node 3"-"node 4"). Therefore, when a failure
occurs in this section, both the "working path 1" and the "working
path 2" are switched over to the backup path to take a detour for
the traffic.
[0160] In the previous cases, a monitoring path has been
established with respect to the common section. However, in this
example, one of the working paths is defined as a representative
path instead of establishing a monitoring path (FIG. 24), and the
error message is transferred using the representative path.
[0161] In this case, when establishing the working path as the
representative path, a representative flag and start-point
information and end-point information of the section, which
indicates that the working path becomes the representative path in
which section, are stored in the PATH message of the
path-establishment signaling protocol (RSVP-TE). For example, the
"node 2" that is a start-point node ID and the "node 4" that is an
end-point node ID are stored as the section in which the working
path becomes the representative path.
[0162] The "node 2" that becomes a path-switch node at the time of
a failure holds the failure-recovery table, as shown in FIG. 25,
which is a table for associating the backup path and the
representative path. A working path ID, a backup-path ID, a
representative-path ID, and a failure section are registered in the
failure-recovery table, in association with each other. In this
example, the path IDs of the "working path 1" and the "working path
2" are "#1" and "#2", respectively, the "backup path 1" and the
"backup path 2" are "#10" and "#20", respectively, and the "working
path 1" is the representative path.
[0163] When a failure occurs between the "node 2"-"node 3"- "node
4", a relay node at an upstream detects the failure. In this
example, there is a failure between the "node 3"- "node 4". Then,
the "node 3" that detected the failure transmits an error message
with respect to the representative path to the "node 2", prior to
other error messages, to notify that the failure has occurred to
the path-switch node "node 2". Alternatively, the "node 3"
transmits the error message with respect to only the representative
path to the "node 2". In the error message, a path ID of the
representative path and an address of a failure point are included
(FIG. 26).
[0164] The "node 2" that received the error message searches the
failure-recovery table, based on the path ID and the address of the
failure point stored in the error message, and identifies the
working path and the backup path. In this case, two sets of the
working path and the backup path are obtained. Then, by
transferring the traffic on the working path obtained to the
corresponding backup path, it is possible to recover from the
failure. As a result, it is possible to recover two working paths
affected from the failure with a reception of the error message
with respect to one representative path (FIG. 27).
[0165] In this manner, by using the representative path, it is
possible to realize a failure recovery of a plurality of paths by
one error message without establishing a new monitoring path.
[0166] As the eighth case, a case in which neither the monitoring
path nor the representative path is used is explained with
reference to FIGS. 28 to 33. First, two working paths are
established by using the path-establishment signaling protocol
(RSVP-TE). One of the two paths is a "working path 1" with a route
of "node 1".fwdarw."node 2".fwdarw."node 3".fwdarw."node
4".fwdarw."node 5"; and the other is a "working path 2" with a
route of "node 8".fwdarw."node 2".fwdarw."node 3".fwdarw."node
4".fwdarw."node 9".
[0167] To recover from a failure occurred at "node 2"-"node
3"-"node 4", a "backup path 1" and a "backup path 2" with a route
"node 2".fwdarw."node 6".fwdarw."node 7".fwdarw."node 4" with the
"node 2" as a start point and the "node 4" as an end point are
established. The "backup path 1" is used as a detour for the
"working path 1", and the "backup path 2" is used as a detour for
the "working path 2". The "backup path 1" and the "backup path 2"
are also established by using the path-establishment signaling
protocol (RSVP-TE), like the working path.
[0168] The "working path 1", the "working path 2", the "backup path
1", "backup path 2" established are shown in FIG. 28. An example of
the label table built with the path-establishment signaling
protocol is also shown in FIG. 28.
[0169] The "working path 1" and the "working path 2" pass the same
section ("node 2"-"node 3"-"node 4"). Therefore, when a failure
occurs in this section, both the "working path 1" and the "working
path 2" are switched over to the backup path to take a detour for
the traffic. In the previous cases, a monitoring path or a
representative path established with respect to the common section
has been used. However, in this example, a single error message is
used for realizing a path switch of a plurality of paths without
using any one of the monitoring path and the representative
path.
[0170] The "node 2" that becomes a path-switch node at the time of
a failure holds the failure-recovery table, as shown in FIG. 29,
which is a table for associating the backup path and a failure
section, to determined a path switch from the working path to the
backup path for which failure (FIG. 30). A working path ID, a
backup-path ID, a representative-path ID, and the failure section
are registered in the failure-recovery table, in association with
each other. In this example, the path IDs of the "working path 1"
and the "working path 2" are "#1" and "#2", respectively, the
"backup path 1" and the "backup path 2" are "#10" and "#20",
respectively.
[0171] When a failure occurs between the "node 2"-"node 3"- "node
4", a relay node at an upstream detects the failure. In this
example, there is a failure between the "node 3"- "node 4". Then,
the "node 3" that detected the failure transmits an error message
with respect to each of the working paths to each start-point node,
to notify that the failure has occurred to the path-switch node
"node 2". In the error message, an address of a failure point is
included (FIG. 31).
[0172] The "node 2" that received the error message carries out the
following process with respect to the error message arrived first.
The "node 2" searches the failure-recovery table, based on the
address of the failure point stored in the error message arrived
first, and identifies the working path and the backup path. In this
case, two sets of the working path and the backup path are
obtained. Then, by transferring the traffic on the working path
obtained to the corresponding backup path, it is possible to
recover from the failure. The same process is carried out for an
error message arrived next, however, when the working path obtained
from a result of the search is already switched over, the
corresponding working path is ignored.
[0173] In the eighth case, because the error message is created for
every working path, a massive amount of error message can be
created at the time of a failure to cause a congestion of the error
message. However, because a route of all of the working paths
affected by the failure is switched based on the error message
arrived first, it is possible to recover from the failure quickly
even when the error messages are created by the congestion (FIG.
32).
[0174] In addition, although the failure-recovery table shown in
FIG. 29 is listed according to a path, by listing the
failure-recovery table as shown in FIG. 33, it is possible to
search for the corresponding working path and backup path at a fast
pace.
[0175] As described above, according to the present embodiment, a
backup path and a monitoring path are set with respect to not only
a plurality of working path sharing the same route but also a
plurality of working paths sharing a part of section, and a
start-point node of the backup path includes the failure-recovery
table 170 in which the working path, the backup path, and the
monitoring path is associated with each other. Upon receiving an
error message, the monitoring-path control unit 180c notifies the
path-switch processing unit 190 that a failure has occurred. The
path-switch processing unit 190 identifies the working path for
which a path switch is necessary and a corresponding backup path.
Then, a process of switching a path from the working path
identified to the backup path is carried out. Therefore, it is
possible to reduce the number of monitoring paths, and to suppress
an increase of a network load at the time of a failure.
[0176] Furthermore, according to the present embodiment, it is
possible to eliminate the monitoring path by using a representative
path instead of the monitoring path. In addition, it is also
possible to eliminate the representative path by carrying out
switching a plurality of paths with a single error message.
[0177] Moreover, according to the present embodiment, although a
case in which the working path control unit 180a, the backup-path
control unit 180b, and the monitoring-path control unit 180c create
the failure-recovery table 170 sequentially when establishing each
of the paths is explained, the present invention is not limited to
this scheme, but can also be applied similarly to a case in which
the failure-recovery table 170 is created regardless of an
establishment of a path.
[0178] According to the present invention, since the number of
working paths that can be handled by a path for a recovery of a
failure, such as a monitoring path, it is possible to decrease the
number of paths for the recovery of the failure.
[0179] Furthermore, according to the present invention, since the
number of working paths that can be handled by a single monitoring
path is increased, it is possible to decrease the number of
monitoring paths.
[0180] Moreover, according to the present invention, since a path
switch from a plurality of working paths is carried out at once, it
is possible to carry out a path switch from the working path
efficiently.
[0181] Furthermore, according to the present invention, since a
monitoring path is identified with ease, it is possible to carry
out a process for the monitoring path efficiently.
[0182] Moreover, according to the present invention, since the
number of working paths that can be handled by a single
representative path is increased, it is possible to decrease the
number of representative paths.
[0183] Furthermore, according to the present invention, since the
number of working paths that can be handled by a single working
path can be increased, it is possible to decrease the number of
paths for a recovery of a failure.
[0184] Although the invention has been described with respect to a
specific embodiment for a complete and clear disclosure, the
appended claims are not to be thus limited but are to be construed
as embodying all modifications and alternative constructions that
may occur to one skilled in the art that fairly fall within the
basic teaching herein set forth.
* * * * *